Epidemiology changes in systemic lupus erythematosus in Italy: the EASIEST study

Abstract

Background

Systemic lupus erythematosus (SLE) shows marked heterogeneity in epidemiology across regions and over time. Understanding temporal changes in incidence, prevalence and patient characteristics is important to interpreting disease burden and assessing the performance of care systems.

Objectives

To evaluate changes after 12 years (2010 vs 2022) in epidemiology, demographic, clinical and therapeutic features of SLE in a general practitioner (GP)-based cohort in Tuscany, Italy.

Methods

We conducted two cross-sectional surveys among GPs in the Florence District (Tuscany, Italy) to identify SLE cases. We assessed temporal variations in incidence and prevalence rates, as well as in sociodemographic, clinical and treatment variables.

Results

SLE incidence doubled from 2010 to 2022, from 5.4 to 10.9 per 100 000 person-years; prevalence increased from 75 to 120 per 100 000 persons. Most diagnoses remained in adults aged 18–60, though non-negligible proportions had paediatric or late onset. Joint, skin and haematological involvement were most frequent, while renal involvement declined from 22% to 16%. Use of biologic therapies grew substantially over time, while glucocorticoids remained pivotal in management. Notably, the proportion of patients not receiving any treatment (glucocorticoids included) increased, indicating a shift towards glucocorticoid-sparing strategies. Furthermore, despite 12 years of demographic change in the general population, the ethnic composition of the SLE cohort remained largely stable, with most patients of Caucasian origin.

Conclusions

Over a decade, this GP-based study shows an increase in SLE prevalence and incidence within participating practices, accompanied by evolving therapeutic strategies and an increasing proportion of patients off treatment, including glucocorticoids. These findings have implications for health planning and support the value of GP-based epidemiologic studies.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Systemic lupus erythematosus (SLE) incidence and prevalence vary widely across Europe and over time, and Italian estimates come from heterogeneous sources (registries, administrative data and referral centres), limiting direct comparability and the evaluation of temporal trends in defined communities.

WHAT THIS STUDY ADDS

• Using the same general practitioner-based approach in the same area 12 years apart, SLE burden increased substantially within participating practices: prevalence rose from 75 to 120 per 100 000, and incidence increased from 5.4 to 10.9 per 100 000 person-years (mean annual incidence 9.24 across 2018–2022).

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These findings quantify a growing community SLE workload and support healthcare planning and resource allocation, while documenting shifts in management (greater biologic use and more patients off treatment at assessment, including glucocorticoids) that should inform future population surveillance and outcomes research.

Introduction

Systemic lupus erythematosus (SLE) is a multisystemic autoimmune disease with a complex pathogenesis, where genetic, environmental and hormonal factors contribute to the immune-dysregulation underlying disease onset.¹ Globally, SLE predominantly affects women of childbearing age, with a female-to-male ratio of approximately 9:1, a data consistently confirmed across different cohorts.²

In Italy, the epidemiological picture has progressively emerged over the last two decades, thanks to several studies. Benucci et al first reported SLE prevalence in the Florence area, providing one of the earliest Italian estimates.³ Shortly thereafter, Govoni et al described incidence and prevalence in the district of Ferrara, confirming values within the lower range of European countries.⁴ More recently, Tsioni et al extended these observations to paediatric populations,⁵ while Zen et al provided a large population-based analysis in the Veneto region, showing an increase in prevalence despite a decline in incidence.⁶ At national level, the Health Improvement Network general practice database analysis offered for the first time a nationwide perspective on huge numbers, corroborating the progressive increase in SLE prevalence.⁷ Recently, the ESCULAPIO study, a large multicentre study conducted across Italian referral centres for SLE care, further highlighted relevant geographic differences, with higher prevalence rates observed in northern Italy.⁸

These Italian data are consistent with global evidence. In Europe, estimates remain heterogeneous, with incidence rates spanning 2 to 23 per 100 000 person-years and prevalence between 30 and 150 per 100 000 persons.⁹ In North America, recent population-based analyses reported SLE incidence ranging from 4.6 to 7.4 per 100 000 person-years and prevalence between 60 and 178 per 100 000 persons, with evidence of decreasing mortality over the last decades.¹⁰ A meta-analysis by Tian et al quantified a worldwide prevalence of 43.7 per 100 000 persons, with marked geographic and sex-related variability.¹¹ Similarly, Barber et al emphasised the growing prevalence of SLE in high-income countries, mainly driven by increased survival and earlier recognition.²

However, incidence and prevalence estimates remain highly variable across countries and over time, reflecting heterogeneity in study design, classification criteria and population structure.¹¹

An important factor influencing comparability among studies is the type of data source used for patient identification (eg, national registries, hospital records, referral centres or general practitioners (GPs)), which may significantly affect case ascertainment.⁹

In this study, we analysed temporal changes in SLE incidence, prevalence, demographic and clinical features and therapeutic management in a well-defined geographic area in Italy between 2010 and 2022, using data provided by GPs.

Materials and methods

Data collection

We conducted two cross-sectional surveys (2010 and 2022) including 10 questions on the clinical and epidemiological characteristics of SLE patients (see online supplemental materials). The questionnaire was distributed via the official mailing list of GPs of the Department of General Medicine of the Florence District. Prevalence was assessed at the two study time points (2010 and 2022), whereas incidence was estimated for each year from 2018 to 2022 to account for potential fluctuations related to the COVID-19 pandemic.

Statistical analysis

Incidence and prevalence rates per 100 000 persons were calculated, with 95% CIs. Incidence was defined as the number of new SLE cases diagnosed per year in relation to the total population followed by participating GPs. The mean annual incidence for the period 2018–2022 was derived from the arithmetic mean of yearly incidences across the interval. Prevalence was estimated as the total number of SLE cases in the population followed by participating GPs. Categorical variables were summarised as frequencies and percentages.

Results

Incidence and prevalence

In 2010, 79 out of 667 (11.8%) contacted GPs participated, covering a patient population of 110 788 individuals. As for the 2022 survey, 84 out of 681 (12.3%) mailed GPs answered, covering 119 093 patients; 83 of them completed all questions of the survey, whereas one GP replied to questions related to SLE incidence and prevalence, but did not respond on sociodemographic, clinical and therapeutic data of his SLE patients (n=2).(online supplemental figure 1).

Among the 79 GPs answering the 2010 survey, 83 SLE cases were reported out of 110 788 patients. Specifically, the estimated SLE incidence and prevalence in the Florence District area resulted in 5.4/100 000 person-year (95% CI 1 to 9) and 75/100 000 persons (95% CI 54 to 86), respectively. For the 2022 survey, the 84 GPs reported 143 SLE cases out of 119 093 patients with an estimated prevalence resulting in 120/100 000 persons (95% CI 94 to 146). The mean annual incidence in the period 2018–2022 was 9.24/100 000 person-years (95% CI 6.3 to 12.5): specifically, the annual incidence was 12.6/100 000 person-year in 2018 (95% CI 11.5 to 13.7) and attested to 10.9/100 000 person-year in 2022 (95% CI 9.7 to 12.1). Detailed results are summarised in table 1 and online supplemental figure 2.

Table 1. Data about SLE incidence and prevalence at different time periods.

	Incidence/100 000 person-year (95% CI)	Prevalence/100 000 persons (95% CI)
2010	5.4 (1.0 to 9.0)	75 (54 to 86)
2018	12.6 (11.5 to 13.7)	–
2019	8.4 (7.0 to 9.8)	–
2020	6.7 (5.2 to 8.2)	–
2021	7.6 (6.2 to 9.0)	–
2022	10.9 (9.7 to 12.1)	120 (94 to 146)
2018–2022	9.24 (6.3 to 12.5)*	–

^*Mean annual incidence during the period 2018-2022

SLE, systemic lupus erythematosus.

Demographic aspects

Sociodemographic variables revealed an increased occurrence of the disease among adult females, with a mean female-to-male ratio ranging between 8.5:1 and 7:1. In most cases, age at onset was between 18 and 60 years; nevertheless, a not negligible proportion of patients experienced a paediatric disease onset (before 18 years) (16.9% in 2010 survey and 11.3% in 2022 survey) or a late disease onset (after 60 years) (24.0% in 2010 survey and 24.1% in 2022 survey).

Most patients were Caucasian (94.0% in 2010 and 92.9% in 2022); the proportion of non-Caucasian patients was 6.0% in 2010 and 7.1% in 2022, with stable ethnic distribution over time (table 2).

Table 2. Demographic features of SLE patients at different time periods.

	2010	2022
SLE patients	83	141
Female	70 (84.0%)	122 (86.5%)
F:M	8.5:1	7:1
Age at onset
Adult onset 18–60 years	49 (59.0%)	91 (64.5%)
Pediatric onset (<18 years)	14 (16.9%)	16 (11.3%)
Late onset (> 60 years)	20 (24.0%)	34 (24.1%)
Ethnicity
Caucasian	78 (94.0%)	131 (92.9%)
American	2 (2.4%)	4 (2.8%)
African	1 (1.2%)	3 (2.1%)
Asian	2 (2.4%)	3 (2.1%)

SLE, systemic lupus erythematosus.

Organ involvement

Data describing organ involvement refer to the overall course of the disease up to the time of questionnaire administration and are summarised in table 3. Articular (63.8% in 2010 survey and 72.3% in 2022 survey), cutaneous (55.4% in 2010 survey and 50.4% in 2022 survey) and haematological (27.7% in 2010 survey and 18.4% in 2022 survey) involvement were the most frequent. Renal disease showed a moderate decrease from 2010 to 2022 as opposed to a slight increase in pulmonary/pleural involvement. Cardiac and central nervous system manifestations remained the less frequent. Data about vascular involvement and severe SLE manifestations were available for the 2018–2022 period only.

Table 3. Prevalence of organ involvement overall and therapies used at survey completion.

	2010 (SLE patients 83)		2022 (SLE patients 141)
	N patients	%	N patients	%
Type of organ involvement
Articular	53	63.8	102	72.3
Cutaneous	46	55.4	71	50.4
Haematological	23	27.7	26	18.4
Pulmonary/pleural	9	10.8	22	15.6
Cardiac/pericardiac	9	10.8	15	10.6
Renal	18	21.6	22	15.6
CNS	3	3.6	10	7.1
Vascular (PE, DVT)	–	–	22	15.6
Severe SLE*	–	–	26	18.4
Therapy†
GC alone	20	24.1	29	20.6
IS‡/antimalarial drugs (± GC)	52	62.6	66	46.8
Biologic drugs§ ± IS (± GC)	4	4.8	25	17.7
No ongoing therapy	7	8.5	29	20.6

^*Nephritis, cerebritis/myelitis, non-infectious pneumonia, infectious mesenteritis, platelet disease

^†Treatment categories are not mutually exclusive; patients may be counted in more than onetreatment category (except for ‘no ongoing therapy’)

^‡Methotrexate, azathioprine, cyclosporin A, mycophenolate mofetil, intravenous immunoglobulins

^§Rituximab or belimumab

CNS, central nervous system; DVT, deep venous thrombosis; GC, glucocorticoid; IS, immunosuppressors; PE, pulmonary embolism; SLE, systemic lupus erythematosus.

Therapy

A specific question of the survey addressed the therapies administered at the time of data collection. A central role for the use of glucocorticoids (GCs) is evidenced by the substantial proportion of patients taking GC monotherapy, which remained stable over time (nearly 20%–25% of patients). At the time of data collection, traditional immunosuppressants and/or antimalarial drugs (± GCs) were reported in 62.6% of patients in 2010 and 46.8% in 2022 (table 3). In parallel, current use of biologic therapies (belimumab or rituximab) increased from 4.8% in 2010 to 17.7% in the later period. The proportion of patients not receiving any therapy at the time of the survey (including GCs) more than doubled, from 8.5% in 2010 to 20.6% in 2022.

Discussion

This study investigated temporal changes in SLE in a well-defined area in Italy, using a survey administered to GPs at two time points separated by more than a decade. We observed a clear increase in both prevalence and incidence between 2010 and 2022. Rising prevalence is generally attributed to prolonged survival and earlier recognition, but the parallel increase in incidence revealed in our study suggests that additional factors may also be involved. Several European cohorts have documented stable or even declining incidence despite rising prevalence,⁶ whereas our results are closer to those from a large nationwide study, using administrative data from a GP-based dbase in Italy, which also reported an increase in both prevalence and incidence.⁷ Interestingly, a recent UK study using the Clinical Practice Research Datalink compared different time periods and likewise found an increase in prevalence accompanied by changes in incidence over time, highlighting how longitudinal analyses with consistent methodology can reveal epidemiological shifts.¹² At global level, Tian et al estimated an SLE prevalence of 43.7 per 100 000 persons and an incidence of 5.4 per 100 000 person-years, with marked geographic variability.¹¹ Together, these findings suggest both an improved case ascertainment in primary care and possibly a true rise in disease occurrence. Overall, our prevalence estimates in 2022 are higher than those reported in previous Italian studies (online supplemental table 1). However, the estimate observed in 2010 (75 per 100 000), obtained using the same methodology in the same geographic area, was closely aligned with existing data, supporting the internal validity of the approach and reinforcing the interpretation of the observed increase over time. In high-income settings, environmental exposures such as air pollution, occupational factors and urban lifestyle have been proposed as contributors to disease onset,² which may also have influenced the trends observed in our area.

The clinical profile of patients remained largely stable over time. Joint, skin and haematological involvement were the most common manifestations, as described in European and North American cohorts.⁹ The only notable change was a modest reduction in renal involvement, from 21.6% to 15.6%, consistent with some evidence of declining lupus nephritis prevalence in recent years, although not consistent across literature.¹³ Improved therapeutic strategies and earlier management may partly explain this finding.

Demographic characteristics were in line with existing literature. The female predominance, with a female-to-male ratio of about 8:1, matches previous data.² Despite demographic changes in the general population, the ethnic distribution of SLE patients remained stable, with Caucasians representing the vast majority in both surveys.

Therapeutic patterns showed continuity, but also important shifts. GCs remained widely used, confirming their central role in SLE management. However, while GCs in combination with conventional immunosuppressants represented most treatments in 2010 (62.6%), this proportion decreased to less than half of patients in 2022 (46.8%). At the same time, the use of GCs alone also slightly declined (from 24.1% to 20.6%), reflecting a gradual transition toward more diversified therapeutic strategies. Conventional immunosuppressants decreased in overall use, in parallel with a marked increase in biologic therapies. Biologics (belimumab and rituximab) rose from 4.8% in 2010 to 17.7% in the later survey, highlighting the progressive integration of targeted treatments into routine practice. Importantly, the proportion of patients off therapy significantly increased over time, more than doubling from 8.5% in 2010 to 20.6% in 2022. This may indicate an increased ability to achieve disease remission or low disease activity and indirectly reflects greater adherence to treat-to-target strategies in patients with SLE.

These data suggest a therapeutic shift characterised by reduced reliance on conventional immunosuppressants, greater use of biologics and a growing proportion of patients who can discontinue chronic treatments, including GCs. This evolution is particularly consistent with the latest EULAR recommendations promoting GCs primarily as bridging therapy and encouraging their tapering and withdrawal when possible.¹

The study has some limitations. Incidence and prevalence were not standardised for demographic variables, which limits comparability with other cohorts. In addition, the response rate of participating GPs was approximately 13% in both study periods, which may limit the representativeness of the estimates. The findings should therefore be interpreted as practice-based sentinel estimates referring to the population covered by participating GPs. The lack of information on non-responding practices precluded assessment of potential differences between responders and non-responders. The absence of practice-level information (eg, practice size, setting or other structural characteristics) did not allow weighting or adjustment of the estimates. Although 40% of the GPs contacted in 2022 were also included among those contacted in 2010, the anonymity of the survey did not allow identification of practices contributing to both periods, and the two surveys should therefore be interpreted as independent cross-sectional assessments. The GP-based design may have introduced misclassification in the absence of centralised validation by a physician expert in the care of patients with SLE. Furthermore, lack of detailed longitudinal therapeutic data precluded analysis of treatment trajectories and long-term outcomes. In addition, the survey design did not allow separate evaluation of antimalarial and immunosuppressive treatments. However, the study assessed the same geographic area more than 10 years apart using the same methodology, which is rarely reported in SLE epidemiology, while the GP-based design ensured broad population coverage, minimising referral bias, and provided insights into real-world management in primary care.